DK168710B1 - Monoclonal antibody or fragment thereof, immunogen, immunoassay method for the determination of glycosylated native human albumin, peptide and immunoassay kit - Google Patents

Description

DK 168710 B1

The present invention relates to a monoclonal antibody or fragment thereof, an immunogen, an immunoassay method for determining glycosylated albumin in a human blood sample using the monoclonal antibody or fragment thereof, a peptide, and an immunoassay kit comprising an antibody of the invention. .

The determination of albumin glycosylation in an individual's blood is a useful measure of glucose concentration control in diabetics.

10 Albumin is the predominant serum protein in blood and has a half-life in the circulation of 10 days. A non-enzymatic glycosylation reaction results in covalent coupling of glucose to a small percentage of albumin molecules in all individuals. Since the non-enzymatic glycosylation rate depends on glucose level of circulation, diabetics with a higher average level have an increase in glycosylated albumin. The severity of the diabetic disorder is therefore reflected in the percentage content of glycosylated albumin.

An analogous reaction takes place between glucose and hemoglobin and produces hemoglobin Alc plus other glycosylated hemoglobins. Hemoglobin has a lifespan of 120 days, and determination of glycosylated hemoglobin values therefore reflects the average circulating glucose level during this period, whereas a glycoalbumin determination represents an average circulating glucose level for 10 days. The importance of glycosylated hemoglobin values has been widely accepted as clinically important for an accurate assessment of the 30 diabetic disorder. It has been relatively easy to develop samples for glycosylated hemoglobin because the hemoglobin molecule is colored and therefore easy to quantify using cheap spectrophotometers. Albumin is colorless, and methods for quantifying glycoalbumin require derivatization of the carbohydrate into a colored product or the protein portion of glycoalbumin to be converted to produce a colored product. Therefore, no large-scale glycoalbumin samples are available in clinical laboratories at present.

A number of glycoalbumin samples have been proposed in the literature. The most important of these are those based on boronate chromatography and thiobarbitur samples. The boronate chromatography method comprises a colorimetric determination of bound protein, e.g. bound to "Glycogel" by Pierce Chemical Co. In this test, serum is applied to a boronate affinity column in which all 10 cis-diol-containing substances (e.g., glycoalbumin and other glycoproteins) are bound. These substances are then eluted and eluate fractions are quantified after the addition of a dye which reacts with proteins producing a colored product. The major drawbacks of this method 15 are that many serum non-albumin proteins are glycoproteins (e.g., immunoglobulins) which are therefore bound and measured by the boronate chromatography sample; the column process comprises several steps for separation and analysis and cannot be readily be automated; and there is data to suggest that glucose interferes with the boronate bond. In the thiobarbitur sample, the ketoamine-protein adduct is converted to 5-hydroxymethylfurfural by hydrolysis with oxalic acid to give a colored product. The major drawbacks here are that hydrolysis requires 2-4 hours at 100 ° C or more; background color must be corrected; and for the moment there are no standards or calibrators.

Glucose's reaction with albumin involves (a) formation of a Schiff's base between glucose C1 and an amino group in albumin and (b) an Amadori rearrangement yielding a 1-deoxyfructosyl carbohydrate covalently coupled to the nitrogen in the amino group. The albumin molecule has 60 potential sites (amino groups) for non-enzymatic glycosylation. This consists of 59 epsilon amino groups in lysine residues and one α-amino group at the 35 N end position of the protein. Of the 60 potential sites, only one lysine is known to be glycosylated in the natural molecule; However, other lysines may be glycosylated to varying degrees. The known lysine has been identified as lysine 525 (the 525th amino acid when counted by the N-terminus of the protein, Garlick et al., J. Biol.

Chem. 258, 6142 (1983)), and the position has been confirmed here. The reason for the specific glycosylation of this lysine and the rapid glycosylation of albumin has not been fully established. The specificity of lysine 525 is likely due to (a) the proximity of an adjacent lysine at position 10 524, thereby lowering the p-α of the one-amino group in lysine 525, making it more reactive in the glycosylation reaction, (b) exposing the lysine 525 side chain to the aqueous exterior of the albumin molecule, and / or (c) the 3-dimensional structure of the albumin molecule which, by an unknown mechanism, increases the reactivity of lysine 525 in the glycosylation reaction.

Monoclonal antibodies have been found to have precise specificity for binding with a variety of organic compounds, including synthetic peptides.

However, despite this technique and a recognized need for an immunoassay for glycoalbumin, there are, so far, no reports that antibodies can be used to obtain glycoalbumin.

The following definitions will be used in the description of amino acid units in peptides.

Amino Acid Abbreviation

Arginine Arg

Aspartic Acid Asp 30 Glutamic Acid Glu

Lysine Lys

Serin Ser

Asparagin Asn

Glutamine Gin 35 Glycine Gly

Prolin Pro DK 168710 B1 4 o

Amino Acid Abbreviation

Threonin Thr

Alanin Ala

Histidine His 5 Cysteine Cys

Methionine Met

Selected Sel

Isoleucine Ile

Leucine Leu 10 Tyrosine Tyr

Phenylalanine Phe

Tryptophan Trp α-Aminobutyric Acid Aba Thus, the object of the invention is to provide a monoclonal antibody specific for glycoalbumin binding, which may serve as a basis for an immunoassay to determine glycoalbumin in human blood samples. There is a widely recognized but so far unmet need for such an immunoassay that would provide a relatively simple and reliable means of assessing the diabetic condition.

In particular, it is the object of the invention to produce a monoclonal antibody that binds specifically to human albumin, which is glycosylated in the lysine residue at position 525, i.e. the 525th amino acid in the peptide chain when counted from the end having the free N-terminal amino group. As used herein, a glycosylated amino acid refers to an amino acid having an amino group which has been modified with a 1-deoxyfructosyl residue by non-enzymatic reaction with glucose.

The present invention fulfills these and other objects and advantages of providing a method for obtaining somatic cell hybridomas which secrete monoclonal antibodies with the desired glyco-albumin specificity. Such hybridomas are formed from conventional fusions between myeloma cells and lymphocytes of an animal, preferably a mouse which has been immunized with an immunogen consisting of a suitable glycosylated peptide chemically linked to an immunogenic carrier. . The glycosylated peptide in the immunogen is obtained synthetically or by proteolysis and comprises, first, a lysine residue whose β, amino group is glycosylated non-enzymatic, and second, at least one other amino acid moiety at a position corresponding to the peptide sequence in human albumin adjacent to the lysine residue at position 525.

The monoclonal antibody or fragment thereof according to the invention is characterized in that it comprises an antibody combining site which specifically binds to an epitope 15 of glycosylated native human albumin, which epitope contains the glycosylated form of the lysine residue at position 525, and the antibody or fragment. thereof, according to a preferred embodiment of the invention, is characterized in that it binds specifically to a glycosylated peptide epitope 20 of the formula

Glyco- (NH) -AA] _- Lys-AA2- wherein Glyco- (NH) is a non-enzymatic glycosylated e-amino group in the lysine residue and one or both of the symbols AA1 and AA2 is a sequence of amino acids , corresponding to the amino acid sequence of 30 human albumin adjacent to the lysine residue at position 525, and if only one of ΑΑχ and AA2 is such a sequence, the other is a bond, a terminal amino or carboxylic group or additional amino acid residues.

Further preferred embodiments of the monoclonal antibody or fragment are apparent from the characterizing parts of claims 3 and 4.

DK 168710 B1 6

In addition to the xnonoclonal antibodies described above, including fragments thereof comprising an antibody combining site, the present invention also includes an immunoassay method for determining glycosylated albumin in a human blood sample such as whole blood, serum or plasma. In the method, the blood sample is contacted with the monoclonal antibody or fragment thereof of the invention. If necessary or desirable, the blood sample is first treated to denature or otherwise expose the epi-peak of lysine 525 in significant amount to any glyco-albumin present in the sample. Then, the binding of the antibody reagent with glycosylated albumin from the sample is determined by any of the usual immunoassay protocols as a function of the glycoalbumin amount of the assayed sample, and the method according to the invention is thus characterized by the characterizing part of claim 6. The present invention also relates to novel and useful peptides and glycosylated forms thereof prepared synthetically or by proteolysis of glycoalbumin or non-20-glycosylated albumin which may serve as a peptide residue in the immunogen of the present invention or which are precursors thereof.

The immunogen of the invention is as claimed in claim 5 and the peptides of the invention are defined in claims 8 and 9.

25 In the drawing, FIG. 1 and 2 are some of the preferred glycosylated peptide fragments or residues which can be bound with a conventional immunogenic carrier to form an immunogen which can be used in the present invention. Structures (1) and (2) show, respectively.

The partial sequence of glycoalbumin in lysine 525 region and the cleavage sites of the proteolytic enzymes trypsin (filled triangles) and V8 protease (blank triangles). The star above lysine 525 shows the site of in vivo glycosylation. Structures (3) - (12) show peptide fragments, 3 ^ which can be synthetically prepared. The stars in addition to lysine 525 show sites of potential additional glycosylation during in vitro synthesis. The sequences are shown in the drawing, and also in the description, from N end position on the left to C end position on the right. Further details and explanations can be found in the examples below.

The monoclonal antibody of the invention is essentially peculiar to its. specificity for binding of the glucosylated peptide sequence in the lysine 525 region of human albumin. This glycosylated residue is the characteristic structural feature of glycoalbumin. An antibody of the invention requires an epitope or determinant site comprising at least the lysine unit modified with 1-deoxy-fructosyl formed after Amadori - rearrangement of the product of the reaction between glucose and the 6-amino group in lysine, as well as a peptide sequence, extending therefrom comprising at least one of the amino acid moieties in the position corresponding to the glycoalbumin sequence adjacent to lysine 525. The other amino acid moieties in the peptide sequence peculiar to the epitope may be the same or different as those which occurs 20 in the natural glycoalbum sequence. In this way, the epitope is peculiar to consist of a carbohydrate and peptide sequence to which the antibody binds, and is unique to the glycolysed lysine 525 sequence in glyco-albumin. Preferably, the antibody is specifically bound to a glycosylated peptide residue of the formula as set forth above, cf. also the characterizing part of claim 2.

Preferably, one or both of AA 1 and AA 2 is a sequence of from 1 to 12 amino acids which corresponds exactly to the amino acid sequence adjacent to the lysine residue at position 525 of glycoalbumin.

The sequence of glycoalbumin comprising 12 amino acids on each side of lysine 525 is as follows (in the direction from N-terminus to C-terminus: Ile-Cys-Thr-Leu-Ser-Glu-Lys-Glu-Arg -Gln-Ile-Lys-Lys (525) - -Gln-Thr-Ala-Leu-Val-Glu-Leu- -Val-Lys-His-Lys-Pro-5 Particularly preferably, the monoclonal antibody of the invention specifically binds with amino acid residues of formula (A) above wherein AA 1 and AA 2 are selected from 10 AA 2: -Lys-Glu-Arg-Gln-Ile-Lys-, -Glu-Arg-Gln-Ile-Lys-, -Arg-Gln- Ile-Lys-, -Gin-Ile, Lys-, -Ile-Lys-,

Lys or a bond, and AA 1: -Gln-Thr-Ala-Leu-Val-G-lu - Gln-Thr-Ala-Leu-Val - Gln-Thr-Ala-Leu-2Q -Gln-Thr -Ala- -Gln-Thr- -Gln-, or a bond.

To produce antibodies according to claim 1, a fragment of the protein chain corresponding to the naturally occurring glycosylated peptide sequence is coupled to a carrier and injected into an experimental animal to elicit an immune response. Lymphocytes, such as spleen cells, from the immunized animal are fused with myeloma cells to produce hybridomas, which are grown and screened for production of monoclonal antibodies. The monoclonal antibodies are screened for those selective to the glycosylated peptide epitope and the particular cell line cloned for use in producing additional amounts of the monoclonal antibody. Review of 35 such monoclonal antibody methods can be found in Melchers et al., Lymphocyte Hybridomas, Springer-Verlag, New York, (1978), 9 DK 168710 B1

Nature 266, 495 (1977), Science 208, 692 (1980), and Methods in Emzymology 73 (Part B) 3-46 (1981).

For the preparation of a suitable immunogen for injection in laboratory animals, e.g. BALB / c mice, rats or the like, either a glycosylated albumin fragment from naturally occurring human albumin or glycoalbumin must be prepared and isolated or chemically synthesized and purified. The glycosylated peptide fragment usable in the present invention, and its non-glycosylated forms or precursors, has the formula (QNH) (NH,) (Cys) - (Tyr) -AA -Lys-AA - (Tyr) s- ( Cys) t (C00H) (B) wherein at least one of the symbols AA 1 and AA 2 is a sequence of from 1 to 12 amino acids corresponding to the peptide sequence adjacent to lysine 525 in human albumin, and if only one of the symbols AA and AA2 is one such sequence, the other is a bond, q and t are each 0 or 1, r and are each 0, 1 or 2, QNH is the β-amino group of lysine, and Q is 1-deoxyfructosyl, and the N-terminal amino group in (NH 2) AA 2 and any lysine units in AA 2 or AA 2 may be glycosylated or non-glycosylated. Preferably, 25 Q is the only glycosylation in the fragment. If one of the symbols q and t is 1, the other is usually 0, and when further q or t is 0, then r or s is also 0.

In a preferred embodiment, glycoalbumin is isolated from human blood and cleaved with a suitable protease enzyme or enzymes to give a convenient size peptide fragments comprising glycosylated lysine 525 and the adjacent amino acids. As albumin is substantially resistant to proteolysis in its natural form, it is usually necessary to de-degrade the protein to the extent necessary for the desired proteolysis to occur. The obtained glycopeptide fragments are isolated by conventional methods such as chromatography on gels with selective affinity for carbohydrate residues and high performance liquid chromatography (HPLC). Preparation of non-glycosylated peptide fragments by cleavage of non-glycosylated albumin with subsequent glycosylation of the fragments is also possible, but clearly much less desirable because of the possibility of glycosylation in sites beyond lysine-525, e.g. other lysine moieties and the N-terminal amino group. The preferred glycosylated peptide fragments produced by proteolysis of human albumin are

QNH

(NH2) Lys-Gln-Thr-Ala-Leu-Val-Glu-Leu-Val-Lys (COOH) and QNH

. (NH 2) Arg-Gln-Ile-Lys-Lys-Gln-Thr-Ala-Leu-Val-Glu (COOH) 2o where Q is 1-deoxyfructosyl. Chemical synthesis can also be used to prepare peptide fragments of the desired sequence by conventional methods and by using peptide synthesis systems on the market. Following peptide synthesis, the peptide obtained is glycosylated under appropriate conditions, e.g. glucose-saturated pyridine or glucose-saturated pyridine / acetic acid = 1: 1 for 48 hours at room temperature. During such in vitro glycosylation, the N-terminal amino group and the β-amino groups of all lysine units of the particular peptide below and in addition to that of lysine 525 can be glycosylated.

Such additional glycosylation may, during immunization, be tolerated differently by the animal which does not respond specifically to such glycosylation due to the distant location or orientation of such glycosylation in the peptide fragment, or it may be selectively removed, e.g. by protease cleavage of one or more terminal amino acids, in particular the potentially glycosylated N-terminal amino acid. The preferred glycosylated and non-glycosylated peptide fragments produced by peptide synthesis are 5 * (QNH) (NH 2) AAg-Gln-Ile-Lys-Lys-Gln-Thr-Ala-Leu-Val - Tyr-AA4 (C00H), * (QNH) (NH2) AA4-Glu-Arg-Gln-Ile-Lys-Lys-Gln-Thr-Ala - Leu (COOH), * (QNH) (NH4) Gln-Ile-Lys-Lys-Gln -Thr-Ala-Leu-Val-Glu - Leu-AA4 (COOH), * (QNH)

15 * I

(NH2) AA5-Gln-Ile-Lys-Lys-Gln-Thr-Ala-Leu-Val- -Glu (COOH) (QNH) (NH2) AAg-Lys-Gln-Thr-Ala-Leu-AA ^ (COOH ), 99 · (QNH) 20 * in (NH 2 AAg-Arg-Gln-Ile-Lys-Lys-Gln-Thr, where Q is 1-deoxyphructosyl, ΑΑ ^ is Lys-Glu-Arg, Arg or a bond , AA 4 is Cys or a bond, AA- is Arg or a bond, i. 4 * Φ AAg is Gln-Ile-Lys, Ile-Lys, Lys or a bond, AA ^ is Tyr-Tyr-Cys, Tyr- Cys, Cys or a bond, AAg is Cys-Tyr-Tyr, Cyr-Tyr, Cys or a bond, wherein the N-terminal amino group and any Lys units in the peptide are glycosylated or non-glycosylated. of the above formulas, Q 1 is deoxyfructosyl and the N-terminal amino group and any other Lys moieties are non-glycosylated.

0 12 DK 168710 B1

A more specific glycosylation of the lysine corresponding to lysine 525 is obtained by combinations of solution and solid phase peptide synthesis coupled with α-amino-blocked β-amino-1-deoxyfructosyl-lysine. In this synthesis, AA0- (Tyr) - (Cys) - (COOH) is prepared by conventional 5 ^ Γ ^ synthesis. Lysine (with ct-amino blocked C-deoxyfructosyl-lysine) is prepared separately and coupled by classical solution-phase chemistry with the amino terminus of such a peptide, giving 10

QNH

(B) -Lys-AA2- (Tyr) g- (Cys) (COOH) where B is a blocking group on the α-amino group of lysine such as tert-butyloxycarbonyl (t-BOC), dinitrophenyl (DNP),

1 O

p-fluorenylmethoxycarbonyl (fMOC) or other suitable blocking groups that can be removed without altering QNH - Lys (1-deoxyfructosyl-lysine). The blocking group can be removed by special chemistry (based on the 20 specific blocking group) which does not alter lysine 1-de-oxyfructosyl structure. This peptide can be used directly as an immunogen or by an immunoassay or can be extended by the addition of NH 2 - (Cys) - (Tyr) -AA

25 QHN

NH - (Cys) - (Tyr) -AA.-Lys-AA «- (Tyr) - (Cys) - (COOH) 2 q £ 1 Z 5 £

The addition may be by a stepwise extension by classical solution or solid-phase peptide synthesis or by a segment condensation where pre-formed NH 2 (Cys) - (Tyr) r ~ AA ^ is condensed on S.

QHN-Lys-AAl- (Tyr) - (Cys) t ~ (COOH) to give the final product.

35 0 13 DK 168710 B1

The introduction of terminal Cys units enables selective coupling of the peptide to immunogenic carriers, such as by bifunctional binding agents known in the art, e.g. m-maleimidobenzoyl-N-sulfosuccinimide stereo (MBS). Alternatively, the peptide fragments may be coupled via the C-terminated carboxyl group by conventional peptide condensation methods, e.g. carbodiimide coupling reagents, Other joining methods known in the art may also be used.

It is generally preferred to introduce one, two or more Tyr units either as terminal units on the peptide or up to the albumin-specific sequence and / or terminal amino acid used for coupling the peptide with immunogenic carriers, e.g. e.g., next to a Cys end device. The presence of Tyr units in the non-specific region of the peptide residue is believed to increase the immunogenicity of the peptide's glycosylated specific region, thereby stimulating the antibody response.

Therefore, in particularly preferred embodiments, AA1 and AA2 in the formulas, respectively. begin with the sequence Cys- (Tyr) - and end with the sequence - (Tyr) -Cys, where r and s are integers from 1 to up to 10 or more, preferably 1 or 2.

The immunogen used to stimulate the production of appropriate immunoglobulins in the most cranial sense will comprise one or more of the glycosylated peptide residues chemically linked with an immunogenic carrier. The general formula for such an immunogen is

Glyco- (NH) carries __r —aa. -Lys-AA - R - - Carrier (C)

m m 1 2 £ O

L Jp where Glyco- (NH), AA ^ and AA2 are as previously defined, provided that AA ^ and AA £ can be terminated amino or carboxyl groups when m or £ is 0, R is a bond or linker group, carrier is an immunogenic carrier, one of the symbols m_ and n ^ is 1 and the other is 0, and £ is on average 1 for the number of available coupling sites on the carrier, and residues AA 1 and AA 2 may comprise Tyr and C devices as discussed above.

The immunogenic carrier material may be selected from those commonly known to have functional groups available for coupling with the glycosylated peptide residue. In most cases, the carrier will be a protein or polypeptide, but other materials may also be used, such as carbohydrates, polysaccharides, lipopolysaccharides, nucleic acids and the like with sufficient size and immunogenicity. For the most part, immunogenic proteins and polypeptides have a molecular weight between 4,000 and. 10,000,000, preferably over 15,000 and, in the rule, over 50,000. As a rule, proteins taken from one animal species will be immunogenic when introduced into the bloodstream of another species. Particularly useful proteins are albumins, globulins, enzymes, hemocyanins, glutelins, proteins with significant non-protein-like constituents and the like. Further discussion of the state of the art regarding conventional immunogenic carriers and methods for coupling hapten can be found at the following sites: Parker, Radioimmunoassay of Biologically Active Compounds, Prentice-Hall (Englewood Cliffs, 30 New Jersey, USA) (1976); Butler, J. Immunol. Meth. 7, 1-24 (1976); Broughton and Strong, Clin. Chem. 22, 726-732 (1976), Playfair et al., Br. Med, Bull. 30, 24-31 (1974) and Weinryb and Shroff, Drug Metab. Rev. 10, 271-283 (1974).

The letter jd in formula C represents the number of glycosylated residues conjugated to the carrier, i.e. the epitopic density of the immunogen, and will range from 0 to the number of coupling sites available on the carrier and may be as high as 5000 in the case of certain high molecular weight synthetic polypeptides such as polylysine. The epitopic density of a particular carrier 5 depends on the molecular guard of the carrier and the density of available coupling sites. Optimal epitopic densities, taking into account convenient and reproducible synthesis of the immunogen and antibody response, are between ca.

10% and approx. 50% of the available coupling groups on the carrier in question.

The linking group R can be virtually any convenient and stable structure. Such a linking group R usually occurs in the form of a bond or an aliphatic chain comprising between 1 and ca. 20 atoms, 15 except hydrogen, and comprising heteroatoms such as nitrogen, oxygen and sulfur. The glycosylated residue may be composed by a variety of groups to form the compound chain R, including methylene, ether, thioether, imino and the like. The skilled artisan has a variety of 20 different linking groups which he can choose from in preparing the immunogen. As a rule, the glycosylated peptide is prepared to end with a functional group such as amino, carboxyl, thiol, hydroxyl or maleimido which are active during a coupling reaction with a suitable group in the carrier molecule.

Particularly preferred immunogens of formula C are those wherein (a) AA 1 is a terminated amino group, AA 2 is Gln-Thr-Ala-Leu-Val-Glu-Leu-Val-Cys, m is 0, and is 1, 30 ( b) AA ^ is (NH₂) Arg-Gln-Ile-Lys, AA2 is Gln-Thr

Ala-Leu-Val-Glu, m is 0, and n is 1, r * r. r 'ψ (c) AA ^ is (NH2) Lys-Glu-Arg-Gln-Ile-Lys, AA2 is Gln-Thr-Ala-Leu-Val-Tyr-Cys, m is 0, and β. is 1, (d) AA ^ is Cys-Glu-Arg-Gln-Ile-Lys, and AA2 is Gln-Thr-Ala-Leu (COOH), m is 1, and n is 0 where Lys is a glycolysed or non-glycolysed lysine unit, B1 (e) AA1 is Gln-Ile-Lys, Ile-Lys, Lys or a terminated amino group, AA2 is Gln-Thr-Ala-Leu-Tyr-Tyr-Cys, in is 0 and n is 1, * \ Λ (f) AA1 is Cys-Tyr-Tyr-Arg-Gln-Ile-Lys, AA2 is c Gln-Thr, m is 1, and n is 0.

The antibody selected for use in an immunoassay may be of any immunoglobulin class, e.g. IgG, IgM, etc. and of any subclass thereof. As a rule, the antibody will be of the IgG class, and if desired, any fragment of such antibody may be used when it merely contains an antibody combining site, e.g. Fab, F (ab ') and Fiab' ^ ·

The antibody reagent of choice may be used in any immunoassay method to determine glycoalbumin in a biological fluid. Such immunoassay methods include the more classical methods such as immunodiffusion, immune electrophoresis, agglutination methods, and complement binding, as well as more current methods including the use of specifically detectable labeling agents such as radioimmunoassays and non-radioisotopic methods. The latter methods can be practiced by a variety of systems, such as competing binding, where a labeled reagent is made to compete with the glycosylated analyte for binding with the antibody reagent. The amount of labeled reagent bound to the antibody reagent or the free type consisting of the labeled reagent which is not thus bound is appropriately measured and can be functionally related to the amount of glycosylated analyte in the sample.

30 In radioimmunoassays, one must be able to physically distinguish between the free type and the bound type, or they must be separable to measure the label, since the signal generated by the label is qualitatively the same for both types. Such a method is known in the art as heterogeneous due to the requirement of phase separation. Other heterogeneous immunoassay methods, including enzyme-labeled immunoassays, are sometimes known, sometimes called ELISA methods (cf. US Pat.

3,654,090) and fluorescent immunoassays (cf. US Patent Nos. 4,201,763, 4,133,639 and 3,992,631), especially fluorescent particle concentration immunoassays 5 (cf. published European Patent Application No. 124,050).

Not too long ago, numerous immunoassay methods have been developed which avoid the separation step using a label whose detectable signal is modulated after the labeled reagent is bound by a binding partner, e.g. antibody. Such methods have become known as homogeneous and can be used advantageously in connection with the present invention because no separations are required and radioisotopes are not involved. Some other methods of this type are fluorescence attenuation and amplification (cf. US Patent No. 4,160,016), energy transfer immunoassay (cf.

U.S. Patent No. 3,996,345) and the immunoassay with double antibody steric hindrance (cf. U.S. Patent Nos. 3,935,074 and 3,998,943). Particularly preferred homogeneous immunoassay methods are those using a label that participates in an enzyme-catalyzed reaction. Examples are substrate-tagged immunoassay (cf.

U.S. Patent No. 4,279,992 and English Patent No. 1,552,607), the immunoassay labeled with a prosthetic group (FAD) (cf. U.S. Patent No. 4,348,565), the enzyme-labeled immunoassay, eg. by means of inhibitor labels (cf. US Patent Nos. 4,134,972 and 4,273,866) and enzyme labeled immunoassay (cf. US Patent 3,817,837).

The novel clonin antibodies of the invention are specific for binding to the glycosylated peptide residue comprising lysine 525 in human albumin. It may in some cases be necessary or it may be particularly desirable to improve the efficiency of analysis of exposing the glycosylated lysine-525 epitope in the natural albumin molecule to perform the desired immunoassay. Controlled access to the epitope can be achieved in any effective manner. Exposure to the epitope in the intact protein is understood to be accomplished by a physical or chemical denaturation or digestion at least 5 in the epitope region. Such denaturation or digestion may be located to the epitope region or may involve a more general or even substantially complete denaturation of the tertiary and further secondary structure of the protein or partial or complete digestion of the protein.

When necessary or desirable, denaturation can occur in a variety of ways, including conventional treatment of the protein by physical means such as heat, sound wave treatment, high or low pH, and, as preferred, chemical denaturation by interaction with a chaotropic agent or chaotrope in solution. Useful chaotropic agents generally include, without limitation, guanidine, urea and various detergents such as sodium dodecyl sulfate (SDS) and others, including the restriction including deoxycholate and certain bile acid salts, 3- (3-cholamidopropyl) dimethylammonio-1 propane sulfonate, organic solvents such as methanol, propanol, acetonitrile and certain salts such as potassium thiocyanate. Nonionic detergents such as "Triton X-100", nonidet-25 NP-40, and octyl glucosides can also act as protein-denaturants. Inclusion of reagents (eg mereaptoethanol or dithiothreitol) which reduce di-sulfide bonds can be effective promoters of the denaturation process. As a rule, protein denaturation can be performed in the most effective manner if combinations of chemical and / or chemical and physical agents (e.g., guanidine and heat, guanidine and SDS, or guanidine and dithiothreitol) are used. Of course, denaturing conditions that result in substantial insolubilization, aggregation or precipitation of the protein should be avoided so that only a negligible amount of the exposed 19 DK 168710 B1

C

epitope is available for resolution for antibody binding. A sufficient amount of denatured protein must be left in solution or suspension to obtain a useful immune bond. The degree of dissolution. The necessary emulsification depends on the circumstances of the intended or desired effect.

The invention will be elucidated hereinafter by means of examples which are not to be construed as limiting 1C.

Example 1

Isolation of naturally occurring albumin and glycoalbumin Whole blood from a normal human donor is collected in a citrate / dextrose solution and separated into red blood cells and plasma by centrifugation. The plasma fraction is chromatographed on a 100 ml boronate / agarose column ("Glycogel", Pierce Chemical Co., Rockford, Illinois, USA) in 0.25 M ammonium acetate, 50 mM MgCl 2, pH 8.0.

The bound fraction (containing mainly glycoalbumin and immunoglobulins) is eluted with 0.1 M Tris, 0.2 M sorbitol and 10 mM EDTA, pH 8.0. To separate glycoalbumin from immunoglobulins, the bound fraction is dialyzed in 50 mM sodium phosphate, pH 8.0, and loaded onto a 100 ml 20 DEAE Affi gel-blue column (Bio-Rad Laboraotires, Rich mond, CA, USA). ). The column is eluted with 50 mM sodium phosphate buffer, pH 8.0, containing a gradient of 0-1.4 M NaCl. The eluate is monitored at 280 nm and the glycoalbumin is identified by SDS-polyacrylamide gel electrophoresis. In some experiments, the plasma is chromatographed on DEA £ - "Affi-gel-blue" before the "Glycogeln column. The separation order does not affect the final purity of the albumin or glycoalbumin.

Non-glycosylated albumin is purified from the unbound fraction of "Glycogel" which is then separated on a DEAS "Affi-gel-blue" column as above. The albumin and glycoalbumin are dialyzed in phosphate buffered saline solution (PBS, 7.2 mM Na 2 HPO 2, 2.8 mM NaH 2 O 2 and 127 mM NaCl, pH 7.4) containing 0.05% sodium azide, lyo -philized and stored at -80 ° C until further use.

5

Example 2

Formation and isolation of enzymatically cleaved glycopeptides from glycoalbumin

Human serum albumin has 59 lysine residues, 10 of which are the primary glycosylation site in the natural molecule. To prepare a peptide containing the glycosylated lysine (at position 525), a known protein sequence in albumin coupled with the unique specificity of the two proteases trypsin and Staphylococcus aureus V8 protease is used. Trypsin cleaves the peptide bond on the carboxy side of lysine and arginine, but does not cleave on the carboxyl side of a glycosylated lysine. Staph. V8 cleaves on the carboxyl side in aspartic and glutamic acid. Therefore, a tryptic digest or V8 digest 20 will produce the respective peptides containing the glycosylated lysine at position 525 (cf. Structures (1) and (2) of the drawing - the filled and blank triangles show the cleavage sites for trypsin and V8, respectively). protease).

In its natural form, albumin is substantially resistant to proteolysis. To optimally expose the protein to the proteases, glycoalbumin is denatured in 8 M urea, 5 mM dithiothreitol (DTT) and 0.1 M ammonium bicarbonate, pH 7.85, for 2 hours at ambient temperature. The denatured protein solution is slowly added to 0.1 M ammonium bicarbonate, pH 7.85, containing a ratio of 1:50 (weight of enzyme: weight of glycoalbumin) protease: protein. The resulting final urea concentration is 0.8 M and the solution is kept at 37 ° C for 16 hours. Then, 35 enzyme (corresponding to weight of the first addition) is added and the solution is incubated for 8 hours. The solution is applied to a boronate "Affi-gel 601" column (BioRad, Richmond, CA, USA) to selectively bind the carbohydrate-containing peptides. The column is carefully washed in 50 mM ammonium bicarbonate and the bound peptides are eluted with 0.1M acetic acid. The eluted peptides are dried, resuspended in 20 mM potassium phosphate, pH 7.0, and injected onto an "Altex-ODS" (4.1 mm x 25 cm) HPLC column (Rainin, Emeryville, CA, USA). A gradient of 1% acetonitrile / minute to a final concentration of 60% acetonitrile in the above buffer 10 is used to elute the bound components. Fractions are collected, dried and hydrolyzed under argon for 24 hours in 6N HCl containing 0.02% phenol. The hydrolysates are dried and analyzed by an OPA pre-column derivatization method [Benson and Hare, Proc. Natl.

15 Acad. Sci. 72, 619-622 (1975)] and separation of amino acid / OPA adducts on HPLC "C-18" column (Supelco, Bellefonte, PA, USA). Amino acids are identified and quantified by comparison with known standards ("Standar H", Pierce Chemical Co., Rockford, Illinois, USA).

The expected and found values for tryptic and V8 peptides are shown in Table I.

25 30 35 DK 168710 B1 0 22

Table I

Staph-V8 Protease Peptide 5

Amino Acid Expected Found GLU 3 3.1 THR 1 1.0 ARG 1 1.2 • / ALA 1 1.0 10 VAL 1 1.0 ILE 1 0.7 LEU 1 1.0 LIGHT 2 1.9 15 Tryptic Peptide

Amino Acid Interest Found.

GLU 2 2.1 THR 1 1.0 20 ALA 1 1.2 VAL 2 1.8 LEU 2 2.0 LIGHT 2 2.) 0 25 30 35 0 23 DK 168710 B1

Example 3

Coupling of proteolytically cleaved glycopeptides with carrier protein

The C-terminated carboxylic acid is selectively activated and coupled with the amines in carrier proteins [Cell 34 587-596 (1983)]. 2 mg of EDCI [1-ethyl-3- (3-dimethylaminopropyl) carbodiimide, Pierce Chemical Co.] in 200 µl Ο, ΟΙΝ HCl is added to 2 mg of the dry glycopeptide of Example 1. This solution is quickly added 6 mg 10 keyhole adhesive hemocyanin (KLH) in 2 ml of water, resulting in a precipitate. The pH of the solution is raised to 9.0 with 0.1 mM ammonium carbonate, stirred for 3 hours and then dialyzed against PBS. The thiobarbituric acid sample [Flucinger and Winterhalter, FEBS Letters 71, 356-360 (1976) 15 shows at least two carbohydrates / molecular weight 100,000 of KLH with fructose as standard.

Example 4

Chemical Synthesis of Albumin Peptides 20 (a) Peptides are synthesized on Applied Biosystems "430A Peptide Synthesizer" (Applied Biosystems, Foster City, CA, USA).

The carboxylene distilled amino acid is coupled to the resin by means of a phenylacetamidomethyl bond 25 having a substitution of 0.7 mmol per g resin. Typically, 0.5 mmol of peptide per synthesis. The N-terminal t-BOC is removed with 60% trifluoroacetic acid (TFA) in dichloromethane (DCM) and the α-amine neutralized with 10% diisopropylethylamine in dimethylformamide (DMF). t-BOC-α-30 mino acids (2 mmol) are converted to preformed symmetrical anhydrides by the addition of 1 mmol of dicylohexylcarbodiimide in 2 ml of dichloromethane. The side chains of the t-BOC amino acids are protected as follows: Arg (TOS), Asp (OBzl),

Cys (4-CH Bzl), Glu (OBzl), His (TOS), Lys (Cl-Z), Ser (Bzl), Thr (Bzl) and Tyr (Br-Z). the t-BOC amino acids Ala, Asn,

Gin, Gly, Ile, Leu, Met, Phe, Pro, Trp and Val are not protected. The dicyclohexylamine salt of t-BOC-L-His (Tos) is converted to the free acid by ion exchange on "AQ-50-X8 (H +)" resin (BioRad) within one hour after coupling. The t-BOC amino acid Asn, Arg and Gln (2 mmol) are coupled by 5 preformed hydroxybenztriazole (HOBt) active esters formed by the addition of 2 mmol HOBt and 2 mmol DCC. The N-terminal t-BOC is removed from the completed peptide and the peptide resin is dried overnight in vacuo.

The protection is completely removed from the peptides which are cleaved from the resin by treatment with anhydrous HF containing 10% anisole for 60 minutes at 0 ° C. The resin is washed with ethyl acetate and the peptide is extracted from the resin with 1.0N acetic acid. The extract is immediately frozen in liquid nitrogen, lyophilized and stored at -20 ° C for later use.

(b) ALB K14C

LYS-GLU-ARG-GLN-ILE-LYS-LYS-GLN-THR-ALA-LEU-VAL-TYR-CYS This peptide of 14 amino acids has an albumin sequence of 12 amino acids with TYR as the penultimate and C-one-distilled CYS . CYS1 sulfhydryl can be selectively coupled with carriers or with fluorescent reagents (see Examples 6 and 9). During in vitro glycolysis, the N-terminal and β-amino groups of the N-terminal lysine are likely to be glycosylated (cf. structure (3) of the drawing). However, proteolytic digestion with V8 will remove terminal LYS. GLU, or digestion with trypsin, will remove terminal LYS-GLU-ARG, resulting in a shorter peptide lacking the potentially glycosylated N-terminated lysine.

(C) ALB Cl IL

CYS-GLU-ARG-GLN-ILE-LYS-LYS-GLN-THR-ALA-Leu

This peptide has 10 amino acids in the albumin sequence plus an N-terminal CYS for selective coupling. The N-terminal amino group in CYS is likely to be glycosylated, but as it is closest to the carrier or label, it should have no effect on the production of antibodies specific for glycarbon albumin [cf. structure (4) of the drawing]. This peptide lacks the C-terminal hydrophobic amino acids VAL and TYR and has greater solubility in aqueous and pyridine buffers.

(d) ALB Q12C

GLN-ILE-LYS-LYS-GLN-THR-ALA-LEU-VAL-GLU LEU CFS

This peptide has 11 amino acids in the albumin sequence plus a C-terminated CYS for selective coupling [cf.

10 of the drawing]. The desired lysine epitope (525) is placed 4 amino acids from the N-terminus, thereby increasing its exposure and antigenicity.

(e) ALB R11E

ARG-GLN-ILE-LYS-LYS-GLN-THR-ALA-LEU-VAL-GLU 15 This peptide has 11 amino acids in the albumin sequence [cf. structure (6) in the drawing]. The N-terminal ARG can be removed by trypsin if the N-one-distilled amino group is glycosylated during in vitro glycosylation. The desired lysine epitope (525) will be 4 amino acids from the N-terminus, thereby increasing its exposure and antigenicity. This peptide also has good solubility in aqueous and pyridine buffers.

(f) ALB Q9C

L & Gln-THR-ALA-LEU-VAL-GLU LEU-VAL-CYS

N-t-BOC-L-Lysine (Sigma Chemical Co., St. Louis, MO, USA) is incubated in 95% pyridine, 5% acetic acid or 10% pyridine in absolute methanol saturated with glucose for 7 days at 50 ° C. The solution is dried to a syrup and purified Nt-BOC-β-1-deoxyfructosyl-lysine by HPLC on a C-18-30 ("Altex ODS", 4.1 mm x 25 cm) column (solvent A = 50 mM triethylamine acetate, pH 6.0; solvent B = A: acetonitrile = 50:50). The product has R 2 of 0.81 on silica gel with chloroform / methanol / acetic acid - 14: 5: 1 and is not reactive with amine-detecting reagents unless exposed to HCl vapors and heated to 100 ° for a short time. C.

0 26 DK 168710 B1 N-t-BOC-L · -1-deoxyfructosyl-lysine is coupled to the N-end position of the synthesized peptide GLN-THR-ALA-LEU - VAL-GLU-LEU-VAL-CYS. This peptide has eight amino acids in the albumin sequence plus a C-terminated CYS for coupling [cf. structure (7) in the drawing]. After removal of t-BOC, the resulting peptide has glycosylated light in 525 exclusively on the β-amino group. For coupling, 1 equivalent of N-t-BOC-β-deoxyfructosyl-lysine in dichloromethane is reacted with 0.5 equivalent of dicyclohexylcarbodiimide for 15 minutes at room temperature under argon. An equal volume of dimethylformamide is added followed by 0.25 equivalent moles of synthesized peptide. After 30 minutes, the solution is dried, resuspended in 25% TFA in dichloromethane for 30 minutes, dried again, and the product is purified by HPLC on a C-28 column.

(g) Tryptic condensation of the glycosylated product of ALB Q9C (AA

The N-terminal extension peptide B-GLN - ILE-LYS is synthesized by conventional solution or solid-solid peptide synthesis. The HPLC purified peptide is incubated with TPCK treated trypsin (Cooper Biomedical, Malvern, PA, USA) in 30% isopropanol or another suitable organic solvent [Fruton, Advances in Enzymology 53, 239-306 (1981)], and equimolar amounts -25 added by LYS-GLN-THR-ALA-LEU-GLU-LEU-VAL-CYS (the product of Example 4 (f)). After 24 hours, the obtained product is isolated B-GLN-ILE-LYS-LYS-GLN-THR-ALA-LEU-VAL - GLU-LEU-VAL-CYS by HPLC. The terminal blocking group (B) can be removed by methods that do not affect the 1-deoxyfructosyl residue on lysine 525.

(h) ALB Q7C

L ^ S-GLN-THR-ALA-LEU-TYR-TYR-CYS

This peptide is used in experiments analogous to those described for (f) ALB Q9C wherein Nt-BOC-35 c-c-1-deoxyfructosyl-lysine is coupled to the N-terminated GLN by dicyclohexylcarbodiimide as described. i 0 27 DK 168710 B1 (f). This peptide (ALB Q7C) has in the C-terminal position a TYR-TYR-CYS structure which is believed to potentiate the immune response against a synthetic peptide immunogen. The small size of the albumin (5 amino acids) of the sequence should give a limited size epitope, thereby focusing the immune response on the glycosylated lysine residue.

(i) ALB K8C

LIGHT-GLN-THR-ALA-LEU-TYR-TYR-CYS

ALB K8C is a small peptide containing the desired glycosylated lysine at the N-terminus and non-albumin TYR-TYR-CYS sequence at the C-terminus. This peptide is readily soluble in 0.1% TFA and absolute methanol, enabling it to be purified by HPLC and glycolysed in the respective solvents.

(J) ALB K9C

LIGHT-LIGHT-GLN-THR-ALA-LEU-TYR-TYR-CYS

The ALB K9C has the characteristics of the ALB K8C plus an additional lysine residue in the N-end position

(k) ALB I10C

20 ILE-LYS-LYS-GLN-THR-ALA-LEU-TYR-TYR-CYS

ALB HOC has (j) the properties of ALB K9C but an additional ILE residue in the N-end position.

(l) ALB Q11C

GLN-ILE-LYS-LYS-GLN-THR-ALA-LEU-TYR-TYR-CYS 25 ALB Q11C has (k) HOC s properties but has an additional GLN residue in the N end position.

(m) ALB ClOT

CYS-TYR-TYR-ARG-GLN-ILE-LYS-LYS-GLN THR

ALB C10T is synthesized with the non-albumin sequence (CYS-TYR-TYR) at the N-terminus to favor antibodies that are preferentially bound to the glycolysed lysine and the C-terminus sequence of this peptide.

35 28 DK 168710 B1 o

Example 5

Glycosylation of synthetic peptides

The four synthetic peptides listed in Table II are glycosylated as listed. The Q9C peptide is prepared by glycosylation as described in Example 4 (f).

Table II

K14C C11L Q12C RllE

(a) Pyridine, 0.25 M glucose + + + + + ίο (b) 50% pyridine, 50% water, 0.125 M glucose + + + (c) PBS, 1.0 M glucose + + + + (d) 95 % pyridine, 5% acetic acid, 0.25 M glucose, pH 7.0 + + + + + 15

The reactions take place from ambient temperature to 50 ° C and for 1-20 days. Samples are dried to a syrup and injected onto an Altex C-18 (1 x 25 cm) using a 0.1% TFA to 0.1% TFA, 60% acetonitrile gradient. Peak fractions are collected, analyzed for carbohydrate and is used in the preparation of glycopeptide - MBS carrier protein immunogens.

The non-glycolysed peptides are also coupled with KLH-MBS and then glycosylated in vitro in PBS containing 1.0 M glucose (pH 9.5 or 7.4) at 37 ° C for 7-14 days. Thiobarbituric acid analysis shows 10-40 carbohydrates / KLH molecular weight 100,000.

Peptides of Example 4 (i) - (m) are glycosylated at elevated temperature (50-80 °) 70 ° C in glucose-saturated methanol for 24 hours. The methanol is removed at negative pressure and the glycopeptide is purified by HPLC. This has been found to be extremely effective in binding glucose with the lysine amino groups of synthetic peptides and with the N-t - BOC-L-Lysine of Example 4 (f).

Glycosylation of N-t-BOC-L-Lysine in glucose-saturated methanol at 50-80 ° C for 24 hours is particularly effective.

0 29 DK 168710 B1

Sequence analysis (Example 6 below) identifies the product (after removal of the α-amine protecting group) as β -deoxyfructosyl lysine. Rapid atomic bomb mass spectroscopy also gives the predicted molecular weight of the glycosylated N-t-BOC-L-Lysine derivative.

Example 6

Sequence analysis to determine the position and amount of glycosylated lysine residues 10 A method for determining the position and amount of glycosylated lysine residues has been developed by automatic Edman gas phase decomposition sequence determination methods. During conventional sequence analysis, both amino groups on lysine are reactive with PITC (phenylisothiocarbamyl), forming a lysine with a PTC group (phenylthiocarbamyl) on the β-amino group and a PTH (phenyl thiohydantoin) on the α-amino group. However, the glycosylated lysine does not want the PTC group of the β, amino group because the carbohydrate blocks this amine from reacting with PITC.

The lysine product is therefore PTH-lysine. By looking at the sequence analysis of the naturally occurring glycoalbumin peptides in Example 2, the PTH-lysine residue has been identified as it has a unique chromatographic retention time on the reverse phase C-18 column used to separate and quantify the various PTH-α mino acids. All glycosylated synthetic peptides are sequenced to identify the particular glycosylated lysine in multi-lysine peptides and to quantify the ratio of lysine to glycolysin. The results show that more than 75% of the lysines have the correct glycosylation reaction product on lysine when the glycosylation is done in methanol.

Example 7 35 Coupling of Synthetic Glycopeptides with Carrier Proteins

Synthetic glycopeptides containing CYS are coupled to carriers as described by Lerner et al. [Proc.

0 30 DK 168710 B1

Natl. Acad. Sci. 78, 3403 (1981)]. Briefly, KLH is reacted with a 50 fold molar excess of sulfo-MBS (Pierce Chemical Co.) for 25 minutes at ambient temperature in 50 mM sodium phosphate, pH 6.2, and the KLH / MBS-5 conjugate is separated from unreacted sulfo-MBS by gel filtration in the same buffer. The KLH / MBS conjugate is immediately added to the dried glydopeptide (2 fold molar excess glycopeptide relative to the maleimide of the carrier) and reacted overnight at room temperature.

10 Por peptides lacking CYS, the C-distilled carboxylic acid is coupled to the amino groups of the carrier proteins as described in Example 3.

Example 8 Immunization

The selected glycopeptide MBS / KLH conjugate of Example 7 is emulsified with an equal volume of Freund's complete adjuvant. Mice (BALB / cBy) are injected with 200 µg of conjugate and boosted after 30 and 60 days with conjugate in 20 incomplete adjuvant. Three days before fusion, mice are injected with 50 µg IV. The mice are sacrificed and their spleen used for fusion according to Kohier and Milstein, Nature 256, 495 (1975). 1 2 3 4 5 6 7 8 9 10 11

Example 9 2

Screening of hybridoma phases for antibodies 3 specific for glyco-albumin 4 (a) ELISA Assay 5

Albumin and glycoalbumin of Example 1 are plated on separate polystyrene microtiter plates (2 µg / ml).

7 100 microliters / well) overnight at 4 ° C. Pla 8 washed with PBS, 0.05% "Tween-20". Upper phases from 9 each cell line are incubated in albumin or glycoalbumin coated plates for 60 minutes. The plates are washed 4 times 11 with PBS + 0.05 "Tween-20" and 200 microliters of secondary antibody are added to each well (1: 2000 dilution 0 of rabbit antimouse IgG peroxidase, Miles Laboratories,

Inc., Elkhart, Indiana, USA). After 60 minutes the plates are washed 4 times in PBS + 0.05% "Tween" and 200 microliters of substrate solution (24.3 mM 5 citric acid, 51.4 mM sodium phosphate, pH 5.3 containing 2.2 mM o phenylenediamine and 5.2 mM hydrogen peroxide). The reaction is terminated after 20 minutes by the addition of 50 microliters of 8 M 2 SO 2 and the product of the peroxidase reaction is read at 492 nm. The monoclonal antibodies specific for glycoalbumin react with glycoalbumin and not albumin.

(b) Fluorescent particle concentration immunolysis

Albumin and glycoalbumin are plated on separate polystyrene particles (Pandes Laboratories, Mundelein, Illinois, USA). Hybridoma phases (20 microliters) are added to each well, followed by 20 microliters of albumin or glycoalbumin coated particles.

After 30 minutes, a 1: 5000 dilution of goat anti-mouse IgG-FITC (cf. Example 10) is added and the incubation is continued for another 30 minutes. All the unbound reactants are removed by filtration and the fluorescence is measured. In a specific response, antibodies in the hybridoma phase are bound with glycoalbumin but not albumin-coated particles.

(c) Fluorescent particle concentration immunoassay with prior dissociation of albumin / glyco-albumin from monoclonal antibody

Goat anti-mouse particles (Pandex Laboratories) 30 are incubated with hybridoma superphases to capture the mouse antibodies. A percentage of mouse monoclonal antibodies is bound to the naturally occurring glycoalbumin in the cell culture media used to grow the hybridoma cells. Unbound components are separated by filtration, leaving the goat anti-mouse particles with bound mouse immunoglobulins from the hybridoma media, and again the glycoalbumin is bound to the mouse immunoglobulin. 20 microliters of 100 mM glycine, pH 3.0, is added to dissociate all complexes. 20 minutes later, 20 microliters of 50 mM Tris base containing glycoalbumin labeled with fluorescein is added by sulfhydryl-specific fluorescein-5-maleimide. . The mouse immunoglobulin / fluorescent glyco-albumin complex is captured by existing goat anti-mouse particles or by the addition of fresh goat anti-mouse Ig particles. The unbound reagents are removed by filtration and the signal is proportional to the mouse anti-glycoalbumin antibodies present in the hybridoma phase.

(d) Improved growth medium

Calf fetal serum (20%) used to maintain and grow myeloma cells and hybridoma cells has a significant concentration of bovine albumin and presumably 20 glycoalbumin which is known to have the same amino acid sequence as human albumin surrounding the glycolysed lysine in position 525. Therefore, it is highly probable that the small amount of anti-glycoalbumin antibodies secreted in the tissue culture medium is immediately bound to the 25 glycosylated albumin molecules and cannot be detected by the standard ELISA assay.

To eliminate the binding of antibody to the glycoalbumin of the medium, myeloma cells and hybridoma cells are adapted for growth in serum and (albumin) free medium.

The medium currently used is commercially available ("HL-1", Ventrex, Portland, Main, USA). The screening of hybridoma superphases in the "HL-1" medium simplifies the identification of clones secreting anti-glycoalbumin antibodies.

(e) Growth of hybridoma cells in glycoalbumin-containing medium

As discussed in Example 9 (d), the presence of glycoalbumin in the medium can prevent the detection of anti-glycalbumin-specific antibodies. Therefore, it is necessary to remove the glycoalbumin from the medium by a selection adsorption process. This is done by selectively absorbing the albumin and glycoalbumin from calf fetal compartments when passing down an "Affi-Gel-blue" column (BioRad Labs, Richmond / CA.USA) using the manufacturer's instructions. The albumin fraction has affinity for the reactive blue dye under these conditions, but 10 can be eluted using 1.4 M sodium chloride. The eluted fraction containing ca. 90% albumin and 10% glycoalbumin are applied to a "Glyco-Gel B" (boronate column, Pierce Chemical Co., Rockford Illinois, USA).

This column selectively binds glycoalbumin. The non-15-bound fraction contains non-glycoalbumin and is again added to the "Affi-Gel" non-bound fraction containing all serum components except albumin. The final mixture is fetal calf serum without glycoalbumin and is used to prepare the medium for growth of hybrids producing anti-glycoalbumin specific antibodies.

Example 10

Fluorescent Labeling of Synthetic Glycopeptides The glycopeptides can be conveniently labeled by sulfhydryl-specific fluorescein conjugates. A two fold excess of fluorescein-5-maleimide in dimethylformamide (40 mg / ml) is added to glycopeptide (10 mg / ml) in 100 mM sodium phosphate, 5 mM EDTA, pH 7.1. Sample 30 is incubated for 20 hours at room temperature. The glycopeptide / fluorescein conjugate is purified by HPLC on an "Altex C-18" 4.1 mm x 25 cm column using 20 mM sodium phosphate to 20 mM sodium phosphate, 50% acetonitrile gradient.

35 DK 168710 Pg 0 34

Example 11

Immunoassay for glycoalbumin in clinical samples

Antibody specific for glycoalbumin is applied to polystyrene particles (0.8 µm) (Pandex Laboratories). The antibody-coated beads (20 microliters) are incubated with an appropriate diluted blood sample, e.g.

1: 800th The sample may be serum, plasma or whole blood. The diluent may be in physiological buffers or denaturing solutions that brighten red blood cells and optimally expose the glycoalbumin epitope. After appropriate incubation (e.g., 5 minutes), synthetic glycopeptide-fluorescein conjugate (Example 10) is added to bind with unoccupied antibody binding sites on the particles. After further incubation (eg 20 minutes).

15, all unbound reactants are removed by filtration and the fluorescein is quantified. Therefore, the fluorescein signal is inversely proportional to the amount of competing glycoalbumin in the clinical trial.

20 25 30 35

Claims (10)

A monoclonal antibody or fragment thereof, characterized in that it comprises an antibody combining site that binds specifically to an epitope in glycosylated native human albumin, which epitope contains the glycosylated form of the lysine residue at position 525. Monoclonal antibody or fragment thereof according to claim 1, characterized in that it specifically binds to a glycosylated peptide epitope of the formula Glyco- (NH) -AA1-Lys-AA2-15 wherein Glyco- (NH) is a non-enzymatic glycosylated e-amino group in the lysine residue, and one or both of the symbols AA1 and AA2 is a sequence of amino acids corresponding to the amino acid sequence of human albumin adjacent to the lysine residue at position 525, and if only one of AA1 and AA2 is a such sequence, the other is a bond, a terminal amino or carboxylic group or additional amino acid residues. Monoclonal antibody or fragment according to claim 1 or 2, characterized in that it binds specifically to the epitope of glycosylated native human albumin which has been treated by physical or chemical denaturation or digestion to expose the epitope for binding. 30 Monoclonal antibody or fragment according to claim 3, characterized in that the epitope is exposed for binding by denaturing with a chaotropic agent. 5. JP wherein Glyco- (NH) is a non-enzymatic glycosylated e-amino group in the lysine residue, one or both symbols AA1 and AA2 is a sequence of amino acids corresponding to the amino acid sequence of human albumin adjacent to the lysine residue at position 525, and if only one of AA 1 and AA 2 is such a sequence, the other is a bond, a terminated amino or carboxyl group or additional amino acid residues, R is a bond or linking group, carrier is a immunogenic carrier material, one of the symbols m and n is 1 and the other is 0 and p is on average 1 for the number of available 20 coupling sites on the carrier. 5. Immunogen of Formula 36 Glyco- (NH) "" Carrier · R -AA. -Ly s-AA ^ —R - * - carries m m 1 J 2 n ji Immunoassay method for determining glycosylated albumin in a human blood sample, characterized in that (a) the blood sample is contacted with a monoclonal antibody or fragment thereof according to claim 1, and (b) the binding of the monoclonal antibody or fragment to glycosylated human albumin is determined as a function of the amount thereof in the sample analyzed. 30 Process according to claim 6, characterized in that the blood sample is treated for physical or chemical denaturation or digestion of glycosylated albumin therein to expose the epitope for binding, preferably by treatment with a chaotropic agent. 0 37 DK 168710 B1 8. Peptide, characterized in that it has the formula (QNH) 5 (NHJ (cys) - (Tyr) -ΑΑ, -Lys-RA - (Tyr) s- (Cys) t {COOH) Z CT in χ * · Wherein at least one of the symbols AA1 and AA2 is a sequence of 1-12 amino acids corresponding to the amino acid sequence, 10 adjacent to the lysine residue at position 525 in human albumin, and if only one of AA1 and AA2 is such sequence, the other is a bond, and the N-terminal amino group of (N ^ JAA ^ and any lysine residue in AA, and AA ~ is optionally glycosylated, and t> each are 0 or 1, r and s, each is 0, 1 or 2, QNH is the β-amino group of Lys and Q is 1-deoxyfructosyl. 9. Peptide, characterized in that it is selected from 25 (a) QNH (NH 2) Lys-Gln-Thr-Ala-Leu-Val-Glu-Leu-Val-Lys (C00H) wherein Q is 1-deoxyfructosyl, 35 0 38 DK 168710 B1 (b) (QNH) in 5 (NH 2) Ar 9 "G 2 -n '" Ile "Iys-Lys-Gln-Thr-Ala-Leu-Val--Glu (COOH) where Q is 1 -deoxyfructosyl, (c) 10 (QNH) * I (NH 1 AA 3 -Gln-Ile-Lys-Lys-Gln-Thr-Ala-Leu-Val - aa AA 1 is Lys-Glu-Arg, Arg or a bond, and AA 4 is Tyr-Cys or a bond, wherein the N-terminal amino group of (NH -20 clay, (d) (QNH) (NH 2) AA 1 -Glu-Arg-Gln-Ile-Lys-Lys-Gln-Thr-Ala -Leu (COOH) Cys or a bond and wherein the N-terminal amino group of (NH 2) AA, and the Lys-3Q moiety of the peptide is optionally glycosylated, (e) (QNH) (NH -Thr-Ala-Leu-Val-Glu-Leu-35 -AA4 (COOH) wherein Q is 1-deoxyfructosyl, AA 4 is Cys or a bond, o g of the N-terminal amino group of (NH 2) Gln and the Lys moiety of the peptide optionally glycosylated, 5 (f) (QNH) * I (NH 2) AA -Val - Glu (COOH) 10 where Q is 1-deoxyrucotsyl, AAj. is Arg or a bond and the N-terminal amino group of (NH 2) AA5 and the Lys moiety of the peptide are optionally glycosylated, (g) (QNH) (NH 2) AA 6 ) Where Q is 1-deoxyfructosyl, AAg is Gln-Ile-Lys, Ile-Lys, Lys or a bond, AA? are Tyr-Tyr-Cys, Tyr-Cys, Cys or a bond, and the N-terminal amino group in (NHO) AA, and 25 x 25 D Lys units in the peptide are optionally glycosylated and (h) (QNH) (NH 2) AAg-Arg-Gln-Ile-Lys-Lys-Gln-Thr 30 where Q is 1-deoxyfructosyl, AAg is Cys. Tyr-Tyr, Cys-Tyr-Cys or a bond, and the N-terminal amino group (N + JAAg and Lys are optionally glycosylated. An immunoassay kit, characterized in that it comprises an antibody according to any one of claims 1-4. 35